[Hamara-devel] discussion about decreasing packaging overhead in debian

[shirish]

Hi all,

This would be a bit longish.

Akshat had come to my place today in the evening and we were talking a 
bit about hamara, about Debian as well as Cent OS. He was of the opinion 
that rpm packaging was much more cleaner, easier than Debian's. While I 
shared with him that in the early days rpm packaging was hard and that 
is one of the reasons why Debian came into play, packaging as well as 
apt and the upgrade scenario.

Now while the positions are reversed, I came just across this on d-d 
mailing list which was discussed few days back.

https://lists.debian.org/debian-devel/2015/11/msg00008.html

It is a whole thread but the interesting part apart from the first 
e-mail was the answer given by David -

https://lists.debian.org/debian-devel/2015/11/msg00014.html

esp. this part -

"Anyway, a giant list of things which could potentially be done isn't
going to change anything as the problem isn't that we have too few tasks
for the giant contributor armies working on the tools which need to be
changed for something to happen. "

So they are open to bringing down the complexity quite a bit but need 
somebody who can take charge of this. If somebody from the community 
would like to take up, would be good. Even if little bit contribution is 
done, this could be turned into a GSOC idea or some other way.

The gist of the issue comes from some small 2 line library as is shared 
by Jeroen Dekkers - 
https://lists.debian.org/debian-devel/2015/11/msg00026.html

Any lesser complexity in Debian packaging would be welcomed by 
everybody, Developers, Maintainers, Uploaders, FTP-Masters and Team, 
Mirrors as lesser data and less things to check but as shared by David 
need somebody who understands this, has skill-sets for this and what is 
not written, is also bit diplomatic and patient as there would be people 
who would want the status quo to remain.


The part about being a bit too late is these kind of ideas and 
discussions are encouraged and supposed to be thought about just after a 
new release happens, the first couple of months because that is when the 
Release goals of the next release are thought about.

https://wiki.debian.org/ReleaseGoals

It isn't necessary or/and probably even possible to have all the Release 
Goals done in new release, as the release goal may depends on some 
specific version of packages (which has that functionality) or specific 
set of circumstances (debconf or such alike where developers can meet 
face to face and have consensus and are able to see the reasons why X or 
Y thinks Z is a good idea) .

We are around half-way for Stretch freeze as the Freeze IIRC was 
announced in October/November 2016.

So while its a great idea, would he be able to get couple of advocates 
and 2-3 volunteers as that is what would be required to push something 
like this which also helps overcome any expected or unexpected push-back 
as well.

@Jonas what is your view on the proposal shared above ?

Any easier packaging process on Debian is also good for Hamara :)

-- 
Regards,
Shirish Agarwal,
Community Lead,
Hamaralinux.org